OSCR

Block oscr:b:7QZGKKD081TM9-O

A block of code linked to a paper, with an identifier OSCR gives it for good: it is never reused nor deleted. What a block identifier is · this block in JSON

Identifier
oscr:b:7QZGKKD081TM9-O
Status
Current: the registry links these lines to the paper today.
Paper
doi:10.1371/journal.pcbi.1009739 · its page here
Code
github.com/nathankong/robustness_primary_visual_cortex, robust_spectrum/loss_functions/InputGradientLoss.py, lines 9 to 91 (InputGradientLoss)
Commit read
0fb7c1b5d945a5e0ed87339300d63b92d63e934b
SWHID
swh:1:cnt:623942e1abe99d4ef18cfc27f8ec3d8f50b84922;origin=https://github.com/nathankong/robustness_primary_visual_cortex;anchor=swh:1:rev:0fb7c1b5d945a5e0ed87339300d63b92d63e934b;path=/robust_spectrum/loss_functions/InputGradientLoss.py;lines=9-91 (the Software Heritage identifier of the file as its authors published it, before any email address is hidden for display; it opens in Software Heritage's archive once the repository is archived there)
In the paper
paragraph 45 (Methods › Convolutional neural network architectures › Robust models); paragraph 42 (Methods › Convolutional neural network architectures › Robust models)
Linked by
the registry's matching program (lexical-v1), which can be wrong
Identifier given
2026-11-16

Open these lines beside the paper

Cite this block

Copy this text: the identifier and its page first, then the code, the paper and the SWHID.

Code block oscr:b:7QZGKKD081TM9-O (https://openscicode.org/b/oscr:b:7QZGKKD081TM9-O)
robust_spectrum/loss_functions/InputGradientLoss.py, lines 9 to 91, in github.com/nathankong/robustness_primary_visual_cortex (https://github.com/nathankong/robustness_primary_visual_cortex), commit 0fb7c1b5d945a5e0ed87339300d63b92d63e934b.
Paper: https://doi.org/10.1371/journal.pcbi.1009739
SWHID: swh:1:cnt:623942e1abe99d4ef18cfc27f8ec3d8f50b84922;origin=https://github.com/nathankong/robustness_primary_visual_cortex;anchor=swh:1:rev:0fb7c1b5d945a5e0ed87339300d63b92d63e934b;path=/robust_spectrum/loss_functions/InputGradientLoss.py;lines=9-91